Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Masters of the Chess: on Chess Capablanca (10 x 8) 40 pieces
1 2 3 4 Previous Next  
Parent - - By neoliminal (*) [us] Date 2007-09-18 16:50
I think this is what you are looking for.  This requirement seems suspect for Gothic Chess, although IANAL:

Nonobviousness Requirement:

If an invention is not exactly the same as prior products or processes (which are referred to as the "prior art"), then it is considered novel. However, in order for an invention to be patentable, it must not only be novel, but it must also be a nonobvious improvement over the prior art. This determination is made by deciding whether the invention sought to be patented would have been obvious "to one of ordinary skill in the art." In other words, the invention is compared to the prior art and a determination is made whether the differences in the new invention would have been obvious to a person having ordinary skill in the type of technology used in the invention.

As can be imagined, the determination of whether a particular change or improvement is "obvious" is one of the most difficult determinations in patent law. In order to make such a determination, an examiner in the patent office will normally review previous patents to find those patents which are closest to the invention in which a patent is sought. If all the features of the invention can be found in a single patent, the examiner will reject the patent as lacking novelty (that is, it is exactly the same as what was previously known and therefore is not new). If no patent contains all of the features, the examiner will attempt to combine two or more prior patents, and attempt to find all of the features in a combination of those prior patents. If the examiner is successful in finding such a combination, the examiner will generally reject the invention as an obvious combination of items known in the prior art. However, there must be some reason to combine the two references, and often a rejection based on such a combination can be overcome.
Parent - By AndrewWalker (**) [au] Date 2007-09-19 14:00 Edited 2007-09-19 14:03
John, this is different, here in Australia and under the PCT (international) rules it is Novelty and Inventive Step. The later of these is not something
you could expect to assess with out training, I still find it tricky in the examining I do (science related). It is determined by making a comparison with prior art
and deciding if there is some amount of "inventiveness" in getting to the new invention.

Method of Manufacture is something else entirely, for all the gory details you can check here: (this is all Australian law) deals with board games and shows they are patentable. I am guessing your last paragraphs come from the US
Patent site. Some of the ways they try to combine two or more prior patents in their reports are "interesting"!
Parent - - By GothicChessInventor (**) [us] Date 2007-09-18 18:04
Hello Andrew,

Australia ranks #3 in terms of Gothic Chess set purchases, right behind the USA and Germany.

Just so you know, in our country, patents are a "necessary evil". If you don't have one, and you create an innovative product, you are "gobbled up" by a bigger company who may not even be in the same field as your business. So, my getting a patent, if businesses want to get involved with you, they need to license your product, and pay you a royalty, typically 7.5%

This is what House of Staunton has done which you can see here:

It also makes sense for those seeking to distribute Gothic Chess products to be a part of some common support structure. If we are all "competing against one another",  that doesn't make too much sense. If we are all promoting one another, that does make sense, and everybody wins.

The patent encourages cooperation, nothing more.
Parent - - By AndrewWalker (**) [au] Date 2007-09-19 14:09
Hi, your comment about "necessary evil" applies in most major countries of the world, but I disagree with the second word!
(especially as it's my work area).

Parent - By GothicChessInventor (**) [us] Date 2007-09-19 20:03 Edited 2007-09-19 20:05
Well, by "evil" I did not mean "evil." Just like good people don't need laws, they are there for the people who would otherwise break them.
Parent - - By jeffniles [us] Date 2007-10-15 01:53
According to this,

The patent expired a year ago, yet Ed Trice is going around continuing to claim the patent. I am no lawyer, but this seems like false patent marking, which could cost him thousands of dollars.

So, the patent is no longer of consequence. Write your Gothic Chess software and sell your Gothic Chess boards. Mr. Trice can do nothing about it now, even though he thinks he can.
Parent - By GothicChessInventor (**) [us] Date 2007-10-15 12:40
Derek Nalls = Jeff Niles.

For a more entertaining read, check out this link:
Parent - - By insipid (**) [us] Date 2007-09-15 18:54
I think computers have gone a long way to refuting the 'draw death' idea.  Super high level computers don't make too many draws when they play each other, they just actually play out the positions.  It's the human side of elite level chess that seems to increase the draw % so dramatically.  When all four round 1 games of the WC are draws before move 30, several with plenty of play left, I don't think the problem is the game but the players.  Sophia rules is all that's really needed IMO.  I'd love to have seen that in the WC, let's play these things out.
Parent - - By JhorAVi (***) [ph] Date 2007-09-16 19:53

> I don't think the problem is the game but the players

When chess was invented it was meant to be played like the way Morphy and Anderseen did in their era.
Our scientific style of play today has outgrown the demands of it's old rules and reveals its unanticipated flaw--  the advent of Draw Death.

Had the inventors of chess knew this would happen i'm sure they would come up with something like gothic :)
Parent - - By GothicChessInventor (**) [us] Date 2007-09-17 01:45
Here is one of my favorite games that ends with some interesting tactics (white just played 21. Qxh5, and don't forget, there is a "j-file" in this game.)

You can replay the whole game here:

Notice in the final position that all 3 of my "SuperMajors" are in some form of jeapordy. The pawn has gxh5 to take the Queen, after which time Kxi6 could take the Chancellor, and Bxf2 can grab the Archbishop at any time.

Black worked out the attacks, and resigned.

I wonder if Gothic Rybka would play 21. Qxh5!!

Parent - - By Gaмßito (****) [cr] Date 2007-09-18 16:18
Hello Ed,

Could you please tell me what is actually the best Gothic Chess program and also what is the best Free Gothic Chess program for download?

Parent - - By GothicChessInventor (**) [us] Date 2007-09-18 17:26

We have a free version of Gothic Vortex available for downloading here:

It comes complete with a 30,000 move opening book and the 3-piece tablebases for Gothic Chess (so it can announce mate in 21 in an ending of Archbishop + King vs. King, etc.)

The only restriction: It never searches more than 7 plies, and it moves instantly.

The full versions of the program don't have this restriction. Even still, most people get hammered by the free version.
Parent - - By Gaмßito (****) [cr] Date 2007-09-19 03:57
Thanks Ed. 
But I was specially looking for a Gotic chess program with no strength restrictions. Full strength is one the most important thing for me in a program. 
Parent - - By George Tsavdaris (****) Date 2007-09-19 12:24
I know about:

-Gothic Vortex(commercial). It can play only Gothic Chess.
Very strong in tactics, compared to the other Gothic Chess programs i will refer. Positionally very bad. It sacrifices material very often to get an attack. Nightmare to play against it! It gets weaker in my opinion in lower time controls or with better words: It gets better as time controls gets longer.

-Smirf(free donationware). It can play all 10x8 and 8x8 Chess and Gothic Chess variants. It does not support directly Gothic Chess but the user for his private use can make it play Gothic Chess with a simple way by loading a PGN that has the Gothic Chess starting setup.
Slightly better positional play from Gothic Vortex but much weaker in tactics. Smirf is (very)weak in (very)short time controls but it gets much better as time increases. Smirf's last version i knew, main problem in my opinion was that in many positionsits behavior was very dumb. It played completely dubious moves frequently enough in a game, while in the next game it will find some nice moves and play great in all the game.
Smirf's general impression in Gothic Chess is that it's a very clever engine(not materialistic many times), but that too frequently loses its mind and plays dumb.

-ChessV(free, as also its source code). It can play many Chess variants. It does not support directly Gothic Chess but the user for his private use, can make it play Gothic Chess with a simple way from version 0.9.2.
ChessV is more steady than Smirf in its play, meaning it does not play so frequently those dubious moves, but generally i find Smirf is more clever. ChessV is strong at tactics from time to time but misses many things, many times.... Positionally is super bad.....

-Zillions of Games(commercial). It can play trillions of games actually and very easily you can make it for your personal use to play Gothic Chess.
ZoG as a general purpose program(it can play almost every type of game you can think of) is weak at Gothic Chess. I win against it every time i play, in every time control. But surprisingly last Smirf's version i had, had problems with ZoG.  ZoG needs time to find some tactics and it's very weak generally at tactics, while positionally is super bad also. It's extremely materialistic.

-XXXXX(private beta). It could play every 10x8 Chess variant with all the pieces(P,K,R,B,N,Q,C,A) so the user for its private use could write at the beginning the Gothic Chess FEN..... The program is discontinued for the last year and it was made by a known Chess programmer of a top Chess engine. I gave as a beta tester, positive results to the programmer but he decided to stop the project and devote to Chess.....
Extremely strong at tactics, much better from Gothic Vortex. Very bad positionally. Materialistic, unlike Vortex, but it could win against Vortex.

Last time i run a Gothic Chess tournament(last year i think) with these programs, i had:
XXXXX:            15 - 1  
Gothic Vortex: 13 - 3
Smirf 1.20:       5 - 11
ZoG:                3.5 - 12.5
ChessV 0.81:   3.5 - 12.5

Both XXXX and GVortex won 4-0 against others while between them XXX won 3-1 against Vortex.
Since i had to play manually i was too lazy to play more games....

Another result i have is XXXX - Gothic Vortex 37.5 -22.5.....

Note that Smirf has evolved since the 1.20 version i used there, as also Gothic Vortex has become much stronger, and ChessV has evolved too.
Gothic Rybka BTW sounds perfect! :)
Parent - By Octopus (**) [de] Date 2007-09-19 14:09 Edited 2007-09-19 14:14
Well, I agree, that SMIRF is better at longer time frames, maybe because it is intended to act rather positional. Nevertheless some bugs have been eliminated since 1.20. Here I post a current testgame PGN between ChessV and SMIRF, using the slower donationware version. Note that in this Embassy Chess game (MBC) "C" means Cardinal (Archbishop/Archangel) and "M" means Marshall (Chancellor/Centaur):

[Event "Embassy Testgames (30 sec / move)"]
[Date "2007.09.19"]
[Time "14:59:57"]
[Round "Game-2"]
[White "ChessV 0.9.2"]
[Black "Smirf BC-168g"]
[Result "0-1"]
[Annotator "RS"]
[SetUp "1"]
[FEN "rnbqkmcbnr/pppppppppp/10/10/10/10/PPPPPPPPPP/RNBQKMCBNR w mKQkq - 0 1"]

{ChessV chrashed finally with: \"assert failed\"} 1. Nc3 Nc6 {(10.22) -0.262} 2.
g3 Nh6 {(10.11) -0.223} 3. Nh3 g6 {(09.07) -0.148} 4. d3 Bxc3+ {(09.01=)
+0.129} 5. bxc3 Cf6 {(10.01) -0.039} 6. Bd2 d5 {(09.01=) +0.156} 7. Rb1 Qd6
{(08.39) +0.139} 8. Nf4 e6 {(08.43) +0.031} 9. Bg2 Md7 {(09.03) -0.010} 10. j4
O-O {(08.39) +0.215} 11. Rb3 a5 {(08.44=) +0.201} 12. e4 dxe4 {(08.13) +0.424}
13. dxe4 a4 {(08.06) +1.008} 14. Rb5 Qa3 {(08.45) +0.812} 15. Ce2 Qxa2 {(08.23)
+0.889} 16. i4 Ne5 {(09.02=) +0.824} 17. i5 Nhg4 {(09.53) -0.012} 18. Nd3 Nxd3+
{(09.04) +0.506} 19. cxd3 c6 {(09.01) +0.596} 20. Rg5 h5 {(09.16) +0.221} 21.
f3 Ne5 {(10.01=) +0.740} 22. d4 Nc4 {(10.06) +0.562} 23. Rc5 Qxd2+ {(09.38)
+0.748} 24. Mxd2 Nxd2 {(10.01=) +0.945} 25. Qxd2 e5 {(10.01=) +0.982} 26. Rj3
Rd8 {(10.02=) +1.049} 27. Rc4 a3 {(10.01=) +1.693} 28. Qa2 Me7 {(10.02=)
+1.697} 29. dxe5 Cxe5 {(11.01=) +2.002} 30. Rc5 Cg7 {(09.10) +2.975} 31. i6
jxi6 {(10.01) +4.049} 32. Qb1 b6 {(09.23) +6.047} 33. Rc4 Ba6 {(10.33) +6.646}
34. Qc1 a2 {(09.49) +6.643} 35. Cf4 a1=Q {(10.19) +5.078} 36. Cxi7+ Kh7
{(12.01=) +6.910} 37. Cj6+ Kg8 {(10.00?) no choice} 38. Qxa1 Bxc4 {(09.31)
+6.990} 39. Qc1 Ch6 {(11.09) +14.89} 40. Ch4 Cxc1 {(12.01=) +15.98} 41. Cxe7+
Kg7 {(10.01=) +M~007} 42. Cf6+ Kxf6 {(08.01=) +M~005 (Mate announcement in
about five moves)} 0-1

Note, that ChessV has big problems (program technically) handling oppressed positions, sometimes even making moves for the opponent's side, which is very strange.
Parent - - By Octopus (**) [de] Date 2007-09-18 21:53
There are some more 10x8 aware free chess programs. In fact both following named programs are multi variant supporting programs with their own GUI. ChessV is moreover an open source project. SMIRF has a free donationware version, which might be (still to be proved) the best 10x8 capablanca piece set supporting free chess program (though still being a beta), moreover supporting X-FEN and PGN import and export. SMIRF is about to be terminated. It might have a successor program named OCTOPUS.

SMIRF download at:
ChessV at:

Parent - - By Gaмßito (****) [cr] Date 2007-09-19 03:38
Thank you very much for your explanation.

I was looking for free Gotic chess programs with no limitations only. I would like to find the strongest program.
I will see how good it is Smirf. Thanks.

Parent - - By Octopus (**) [de] Date 2007-09-19 10:07 Edited 2007-09-19 10:10
Gothic Chess is a USA patented game. Therefore you probably will not find any FREE and UNRESTRICTED Gothic Chess supporting chess program. But there are a lot of other interesting free 10x8 board setups (partially with different castling rules) like: Janus Chess, Embassy Chess (MBC), Optimized Chess and of course Capablanca Random Chess (CRC). Some programs like SMIRF or CHESSV have released special Gothic Chess supporting internal / private releases for to become able to participate in GC tournaments like the one in end of this year 2007. 

Parent - - By GothicChessInventor (**) [us] Date 2007-09-19 16:54
So far for the 2007 Tournament we have the following Gothic Chess programs:

1. Gothic Vortex
3. TSCG Gothic 64 (first 64 bit Gothic Chess Engine)
4. Tornado (new program I never heard of before)
5. ChessV

I am wondering if this "Tornado" program is the same one George is speaking about?
Parent - - By George Tsavdaris (****) Date 2007-09-19 21:18

>I am wondering if this "Tornado" program is the same one George is speaking about?

I don't know. What is the programmer's name of Tornado?
If he is a programmer of a top Chess engine then i guess he is the one..... :)
As you know the programmer has asked from me not to give his name and even refused to have a communication with you, when you asked me to put him in touch with you by email.... In fact he was a bit annoyed with me because i gave results of his program to the public.....

I have to hear from him for about 8 months so....

I hope he is the one, since it was a pity that the development of that great program had stopped.

TSCG is, i suppose, a modified by Michel Langeveld engine derived from TSCP?
Last time was strong enough. I hope it is again now.
Too bad it is private.... :(
Since it is derived from a free source code Chess program can we persuade Langeveld to provide the program for downloading? Or even better the source code? :)
Parent - - By GothicChessInventor (**) [us] Date 2007-09-20 19:34
Michel's website  is here:

Maybe you can talk him into a public source code release. His program will be the only 64-bit engine in this tournament, unless Tornado is also.
Parent - By AndrewWalker (**) [au] Date 2007-10-17 11:17
If Michel ever reads this, there is no such thing as an international patent as mentioned on the site.

Parent - - By Octopus (**) [de] Date 2007-09-20 08:47
On Tornado: is this a German program from Mainz?
Parent - By GothicChessInventor (**) [us] Date 2007-09-20 19:31
The Tornado program is apparently from England. The programmer would not disclose his name and refers to himself as "Colin".
Parent - - By jeffniles [us] Date 2007-10-15 01:58
It is no longer patented according to this,

So, perhaps some free implementations will arise, now that Mr. Trice can no longer stop anyone from writing them.
Parent - By GothicChessInventor (**) [us] Date 2007-10-15 12:02 Edited 2007-10-15 12:49
When I was in Iceland in 2006 to negotiate terms for the Fischer-Karpov Gothic Chess match, and when another patent dispute was ongoing at around the same time (see Harac vs. Trice for more info), I signed a Power of Attorney transfer of the rights over to (you guessed it) my attorney. The fees were paid in my absence under the direction of his firm. The USPTO typically scans for checks written to them by patent owners, and in the absence of a direct match against a crossreference check of the patent numbers, the list you posted gets kicked  out. It's by no mean comprehensive nor official. Mostly it's a means to get those on the list to contact them and say: "Hey, what the hell.... I paid you guys..." and the next iteration of the list comes out, etc., and so on.

Since their information technology is not very intricate, they have no means to incorporate Power of Attorney transfers into this schema, and since there was no match for "Trice" and "6481716", they made a mistake. There is no way for them to "manually" check this list, they assume those that don't write back were the ones who did not renew their patent(s). You can make all the niose you want, I'll be there to collect my royalty at the end of the day.

Incidentally, I was told by a colleague that the General Accounting Office indicates that the USPTO has at least 10% of their disseminated information in error. That's what, like an A- if a teacher assigned a grade to it?

And, by the way, the licensing fee was only $1 per year, I don't think any INDIVIDUAL was scared away by that. Unless you are like really really poor. The patent is primarily to deal with corporate-to-corporate negoations, like the various QVC deals, Hasbro, and if Microsoft ever wanted to make a really crappy program, etc.
Parent - By GothicChessInventor (**) [us] Date 2007-09-19 16:55

Why don't you play the free version of Gothic Vortex and let us know how you do. I think you will be pleasantly surprised at its strength, even in blitz.
Parent - By lkaufman (*****) Date 2007-09-17 05:01
Although I favor the "Sophia" rules, I think they only make a small difference; most of the games that are agreed draws in these super tournaments would still end in draws if played to the end. When top level programs play each other in standard length games, the draw percentage is at least as high or higher than in top human events, provided that the programs are of about the same strength (which doesn't apply when Rybka is on one side and not the other!). When Rybka plays an older (weaker) version of herself (substantially different), the draw percentage climbs steadily with depth, passing 50% at a level much faster than game in one minute on my quad. So I'm afraid I can't agree with your conclusion.
Parent - - By neoliminal (*) [us] Date 2007-09-16 17:09
I've been asked to post the following piece values for 10 x 8 boards.  Please comment on them:

Q = 9 (9.3) 344
A = 9 (9) 333
M = 8 (7.76) 287
R = 5 (4.84) 179
B = 3 (3.05) 113
N = 3 (3.27) 121
P = 1 (1.14) 42

I've been asked not to disclose the system used to created these values.
Parent - - By GothicChessInventor (**) [us] Date 2007-09-16 17:14
People might not know what "A" and "M" represent. Also, if you are reporting 3 different values for each piece, you should label them better. Such as:

Group 1

Queen = 9
Archbishop (A piece that can move like a Knight or Bishop) = X
Chancellor (A piece that can move like a Knight or Rook) = Y


Group 2

Queen = 9.3

Parent - - By neoliminal (*) [us] Date 2007-09-16 17:52
Thanks Ed.  I was assuming people were familiar with the Casablanca set.
Parent - - By GothicChessInventor (**) [us] Date 2007-09-16 19:30
Casablanca was a movie starring Humphrey Bogart.
Parent - - By neoliminal (*) [us] Date 2007-09-17 00:15
Too true.  "...I'm going to miss you. Apparently you're the only one in Casablanca with less scruples than I. "
Parent - - By GothicChessInventor (**) [us] Date 2007-09-17 05:03
The correct grammar is "fewer", not "less", but I am not familiar enough with the dialog at the end of the movie where Bogart is walking away from the small airstrip to remember if that was actually said.

One quote I do remember:

"I am shocked, I say, shocked to learn there is gambling going on in this establishment."

"Your winnings sir."

"Thank you."
Parent - By neoliminal (*) [us] Date 2007-09-18 00:10
I assure you he says 'less.'
Parent - By Sesse (****) [no] Date 2007-09-17 07:10
And Ingrid Bergman! Sheesh. :-)

/* Steinar */
Parent - - By Sesse (****) [no] Date 2007-09-16 19:53
You still have not explained what the actual numbers represent, in particular the rightmost column.

/* Steinar */
Parent - By neoliminal (*) [us] Date 2007-09-17 00:12
Just ignore the third column.  It's simply a multiple of the second column and the first column is simply a rounding for the second.
Parent - - By lkaufman (*****) Date 2007-09-17 01:57
I'm no expert on Gothic or Capablanca chess, but I know that the Archbishop is not as powerful as the Chancellor (apparently M in this table) and is not even close in value to the queen, so the "A" value is way too high and the "M" value somewhat too low. As in regular chess, the bishop's value depends greatly on whether the bishop pair is scored separately or not; if not (as I assume since it's not in the table) the bishop's value is too low.
Parent - - By GothicChessInventor (**) [us] Date 2007-09-17 03:09
I interviewed Larry for one issue of Gothic Chess Review in the year 2001. I asked him about some of the weights for the new pieces. With his permission, I will post his values.
Parent - - By lkaufman (*****) Date 2007-09-17 04:51
Sure, though I don't remember what I said. I hope I don't contradict myself here!
Parent - By GothicChessInventor (**) [us] Date 2007-09-17 06:50
This is a summary of my original mathematical research, for those who seek a rigorous approach to trying to assign weights to the pieces on the 80-square board:

Of course those numbers are by no means final, but they seem to be doing Gothic Vortex a great deal of good.

And some quotes from Larry's orginal interview in the January 2001 issue of Gothic Chess Review:

"The Chancellor should be worth about 1/2 to 3/4ths of a pawn or so less than the Queen... The difference between the Queen and the Archbishop should be less than the difference between a Rook and a Knight (about 1.75 pawns in chess) because adding the movement of the Bishop cures the Knight's limited range. This implies that the Chancellor is worth about 3/4ths of a pawn or so more than the Archbishop."
Parent - - By neoliminal (*) [us] Date 2007-09-17 21:48
Could you expound on your thoughts regarding the valuation of the Archbishop, Chancellor and Queen vis-a-vis their relative values.  For example, would combining two early pieces in the Archbishop give it a higher early valuation than the Chancellor  which combines an early and end game piece?  Also, take into consideration that the Archbishop is the only piece that can administer mate on it's own.

Thank you.
Parent - - By GothicChessInventor (**) [us] Date 2007-09-17 22:17
Well, it depends what you mean by "on its own." From the Trice-Andreas Kaufman game (no relation to Larry) if black plays 21..gxh5 then 22. Ch8# is mate, "all on its own" the Chancellor is giving mate.
Parent - - By neoliminal (*) [us] Date 2007-09-18 13:57
"The new pieces have properties that enrich the game. For example, the archbishop can checkmate a lone king by itself (king in a corner, archbishop placed diagonally with one square in between)."  Wikipedia entry on Capablanca.
Parent - By GothicChessInventor (**) [us] Date 2007-09-20 03:58
A fact that has been well known since this wild 3 minute game played on August 10, 2000
Parent - - By neoliminal (*) [us] Date 2007-09-18 13:39
I have several issues with the values give for the Gothic Chess ( system, in particular the method of valuation.

1. I think the devaluation of the knight based on a safe check (Taylor, 1876) system is not applicable when the board is still 8 squares deep.  It would be more applicable if the board were 10 squares deep but the width of the board does not significantly impact the value of the knight.  However the positioning of the knight is key.  Obviously if the knight were placed in the corner of the board at the start of the game, the value of the knight would be significantly lower.  Since the Archbishop is centrally located this increased it's value in Gothic Chess.

2. The safe check system for valuation is suspect.  Based on the values given for Gothic Chess a larger board devalues the rook.  This is clearly wrong.  Slider values increase on larger boards.  Imagine a board 10 x 100.  The rook, in this scenario, would be valued extremely lowly relative to the standard 8x8 board.  Trice makes the error of comparing them thusly:
"Comparing 1/6 = 0.16666666667 to 0.15759493671 indicates that a Rook on the 10x8 Gothic Chess board is about 94.5% of the strength of the Rook on an 8x8 board. So, the Gothic Chess Rook is about 5% weaker than its chess counterpart."  On a board 10x100 the value of a rook would then be even lower... but the ability to move 99 squares is incredible powerful on a board 10x100.  If anything, the value of the rook should be increasing as the board size increases.

3. The safe check system ignores blocking pieces and the ability of the knight to attack over interposed squares.  This ignorance results in the 16% lose of value that is estimated by Trice in his use of the system.

The safe check system, while useful in it's time period (1800s) is not a modern system of evaluation of pieces.
Parent - - By GothicChessInventor (**) [us] Date 2007-09-18 17:36
There is a major flaw in your thinking. The strength of a piece is proportional to the number of squares it can reach (a linear, scalar multiple of a "power-of-one" variable set) divided by the area of a board it resides on (a "squared" variable, board length x board height.)

If what you are saying is true, Rooks would be weaker on a board of dimension 5x5 because they could only reach 8 squares, yet these 8 squares are almost 1/3rd of the whole board (8/25). Imagine how strong a Rook would be on a 64 square board if it could reach 20 squares (8/25 x 64). It would be almost as strong a Queen.

One way to measure piece strength would be to look at the "longest wins" in the tablebases. I computed the complete 5-piece set, some of the endings are shown here:

You can see that on a 10x8 board, the longest win for King + Queen + Pawn vs. King + Queen is 268 moves. On the 8x8 board, is it about 120 moves.

If the Queen is stronger on a 10x8 board than on an 8x8 board, why does it take twice as long to mate from the longest win? Shouldn't the win lengths decrease if piece strength increases?

Expain that.
Parent - - By neoliminal (*) [us] Date 2007-09-18 19:17 Edited 2007-09-18 19:20
Sure.  The difference is that you let the computer turn the pawn into a knight as a possible option.

Put a Queen + Rook + King vs. King + Rook.  How much longer does that take on a 10x8 vs 8x8?
Parent - - By GothicChessInventor (**) [us] Date 2007-09-18 19:27
You're an idiot.

The "longest win" is the most difficult position to win, not a winning sequence with deliberately non-optimal moves.

The fastest win in that position on the 10x8 board requires 268 moves.

The fastest win in 8x8 chess for longest winning position involving Queen + Pawn vs. Queen is about 120 moves.

So why is the queen stronger on a 10x8 board if it takes over twice as long to win?

I'll answer that for you, since it's beyond your comprehension:

The queen is weaker on a 10x8 board than on an 8x8 board.
Up Topic Rybka Support & Discussion / Rybka Discussion / Masters of the Chess: on Chess Capablanca (10 x 8) 40 pieces
1 2 3 4 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill